If PTP_1588_CLOCK is n, building fails:
drivers/net/ethernet/aquantia/atlantic/aq_ptp.c: In function aq_ptp_adjfine:
drivers/net/ethernet/aquantia/atlantic/aq_ptp.c:279:11:
error: implicit declaration of function scaled_ppm_to_ppb [-Werror=implicit-function-declaration]
scaled_ppm_to_ppb(scaled_ppm));
Just cp scaled_ppm_to_ppb() from ptp_clock.c to fix this.
Fixes: 910479a9f793 ("net: aquantia: add basic ptp_clock callbacks")
Signed-off-by: YueHaibing <[email protected]>
---
drivers/net/ethernet/aquantia/atlantic/aq_ptp.c | 20 ++++++++++++++++++++
1 file changed, 20 insertions(+)
diff --git a/drivers/net/ethernet/aquantia/atlantic/aq_ptp.c b/drivers/net/ethernet/aquantia/atlantic/aq_ptp.c
index 3ec0841..80c001d 100644
--- a/drivers/net/ethernet/aquantia/atlantic/aq_ptp.c
+++ b/drivers/net/ethernet/aquantia/atlantic/aq_ptp.c
@@ -262,6 +262,26 @@ static void aq_ptp_tx_timeout_check(struct aq_ptp_s *aq_ptp)
}
}
+static s32 scaled_ppm_to_ppb(long ppm)
+{
+ /*
+ * The 'freq' field in the 'struct timex' is in parts per
+ * million, but with a 16 bit binary fractional field.
+ *
+ * We want to calculate
+ *
+ * ppb = scaled_ppm * 1000 / 2^16
+ *
+ * which simplifies to
+ *
+ * ppb = scaled_ppm * 125 / 2^13
+ */
+ s64 ppb = 1 + ppm;
+ ppb *= 125;
+ ppb >>= 13;
+ return (s32) ppb;
+}
+
/* aq_ptp_adjfine
* @ptp: the ptp clock structure
* @ppb: parts per billion adjustment from base
--
2.7.4
> If PTP_1588_CLOCK is n, building fails:
>
> drivers/net/ethernet/aquantia/atlantic/aq_ptp.c: In function aq_ptp_adjfine:
> drivers/net/ethernet/aquantia/atlantic/aq_ptp.c:279:11:
> error: implicit declaration of function scaled_ppm_to_ppb [-Werror=implicit-function-declaration]
> scaled_ppm_to_ppb(scaled_ppm));
Hi Yue,
Thanks for noticing this. It seems I've added scaled_ppm_to_ppb usage but did
not checked PTP_1588_CLOCK=n case after that.
>
> Just cp scaled_ppm_to_ppb() from ptp_clock.c to fix this.
I'm honestly not sure if duplicating the code is a good way here.
I'll think on how to exclude at_ptp at all on such a config.
Regards,
Igor
On Fri, Oct 25, 2019 at 09:37:26PM +0800, YueHaibing wrote:
> If PTP_1588_CLOCK is n, building fails:
>
> drivers/net/ethernet/aquantia/atlantic/aq_ptp.c: In function aq_ptp_adjfine:
> drivers/net/ethernet/aquantia/atlantic/aq_ptp.c:279:11:
> error: implicit declaration of function scaled_ppm_to_ppb [-Werror=implicit-function-declaration]
> scaled_ppm_to_ppb(scaled_ppm));
>
> Just cp scaled_ppm_to_ppb() from ptp_clock.c to fix this.
>
> Fixes: 910479a9f793 ("net: aquantia: add basic ptp_clock callbacks")
> Signed-off-by: YueHaibing <[email protected]>
Hi YueHaibing,
thanks for your patch.
What is the motivation for not using the existing copy of this function?
> ---
> drivers/net/ethernet/aquantia/atlantic/aq_ptp.c | 20 ++++++++++++++++++++
> 1 file changed, 20 insertions(+)
>
> diff --git a/drivers/net/ethernet/aquantia/atlantic/aq_ptp.c b/drivers/net/ethernet/aquantia/atlantic/aq_ptp.c
> index 3ec0841..80c001d 100644
> --- a/drivers/net/ethernet/aquantia/atlantic/aq_ptp.c
> +++ b/drivers/net/ethernet/aquantia/atlantic/aq_ptp.c
> @@ -262,6 +262,26 @@ static void aq_ptp_tx_timeout_check(struct aq_ptp_s *aq_ptp)
> }
> }
>
> +static s32 scaled_ppm_to_ppb(long ppm)
> +{
> + /*
> + * The 'freq' field in the 'struct timex' is in parts per
> + * million, but with a 16 bit binary fractional field.
> + *
> + * We want to calculate
> + *
> + * ppb = scaled_ppm * 1000 / 2^16
> + *
> + * which simplifies to
> + *
> + * ppb = scaled_ppm * 125 / 2^13
> + */
> + s64 ppb = 1 + ppm;
> + ppb *= 125;
> + ppb >>= 13;
> + return (s32) ppb;
> +}
> +
> /* aq_ptp_adjfine
> * @ptp: the ptp clock structure
> * @ppb: parts per billion adjustment from base
> --
> 2.7.4
>
>
On 2019/10/26 16:09, Simon Horman wrote:
> On Fri, Oct 25, 2019 at 09:37:26PM +0800, YueHaibing wrote:
>> If PTP_1588_CLOCK is n, building fails:
>>
>> drivers/net/ethernet/aquantia/atlantic/aq_ptp.c: In function aq_ptp_adjfine:
>> drivers/net/ethernet/aquantia/atlantic/aq_ptp.c:279:11:
>> error: implicit declaration of function scaled_ppm_to_ppb [-Werror=implicit-function-declaration]
>> scaled_ppm_to_ppb(scaled_ppm));
>>
>> Just cp scaled_ppm_to_ppb() from ptp_clock.c to fix this.
>>
>> Fixes: 910479a9f793 ("net: aquantia: add basic ptp_clock callbacks")
>> Signed-off-by: YueHaibing <[email protected]>
>
> Hi YueHaibing,
>
> thanks for your patch.
>
> What is the motivation for not using the existing copy of this function?
I'm not sure if PTP_1588_CLOCK is needed at this config,
using the existing function need to PTP_1588_CLOCK is selected.
>
>> ---
>> drivers/net/ethernet/aquantia/atlantic/aq_ptp.c | 20 ++++++++++++++++++++
>> 1 file changed, 20 insertions(+)
>>
>> diff --git a/drivers/net/ethernet/aquantia/atlantic/aq_ptp.c b/drivers/net/ethernet/aquantia/atlantic/aq_ptp.c
>> index 3ec0841..80c001d 100644
>> --- a/drivers/net/ethernet/aquantia/atlantic/aq_ptp.c
>> +++ b/drivers/net/ethernet/aquantia/atlantic/aq_ptp.c
>> @@ -262,6 +262,26 @@ static void aq_ptp_tx_timeout_check(struct aq_ptp_s *aq_ptp)
>> }
>> }
>>
>> +static s32 scaled_ppm_to_ppb(long ppm)
>> +{
>> + /*
>> + * The 'freq' field in the 'struct timex' is in parts per
>> + * million, but with a 16 bit binary fractional field.
>> + *
>> + * We want to calculate
>> + *
>> + * ppb = scaled_ppm * 1000 / 2^16
>> + *
>> + * which simplifies to
>> + *
>> + * ppb = scaled_ppm * 125 / 2^13
>> + */
>> + s64 ppb = 1 + ppm;
>> + ppb *= 125;
>> + ppb >>= 13;
>> + return (s32) ppb;
>> +}
>> +
>> /* aq_ptp_adjfine
>> * @ptp: the ptp clock structure
>> * @ppb: parts per billion adjustment from base
>> --
>> 2.7.4
>>
>>
>
> .
>
>> Hi YueHaibing,
>>
>> thanks for your patch.
>>
>> What is the motivation for not using the existing copy of this function?
>
> I'm not sure if PTP_1588_CLOCK is needed at this config,
> using the existing function need to PTP_1588_CLOCK is selected.
Hi YueHaibing,
Please checkout this patch: https://patchwork.ozlabs.org/patch/1184620/
It fixes the problem without duplicating the function.
Regards,
Igor
On 2019/10/26 19:09, Igor Russkikh wrote:
>
>>> Hi YueHaibing,
>>>
>>> thanks for your patch.
>>>
>>> What is the motivation for not using the existing copy of this function?
>>
>> I'm not sure if PTP_1588_CLOCK is needed at this config,
>> using the existing function need to PTP_1588_CLOCK is selected.
>
> Hi YueHaibing,
>
> Please checkout this patch: https://patchwork.ozlabs.org/patch/1184620/
>
> It fixes the problem without duplicating the function.
Yes, this fix the issue.
>
> Regards,
> Igor
>
>